Closed benbovy closed 6 years ago
Hi Benoit, this is great! Thanks for the changes.
How did the benchmark go on your machine?
Thanks @benbovy for this work!
How did the benchmark go on your machine?
Not bad I would say, although I'm not much experienced yet to have good interpretation of these benchmarks (I was just curious :-) )
Here are the results (Macbook Pro 15'' Mid 2015, 2.5 GHz i7):
I just wanted to run the benchmarks on my machine (MacOS) but I had to fix the following:
In
environment.yml
: use the conda-forge channel and also install cmake, pythran dependencies and pythran via pip (no conda package available for osx on the "serge-sans-paille" conda channel).In
CMakeLists.txt
: download googlebenchmark before adding the googlebenchmark sub-directories.In
CMakeLists.txt
: instead offind_package(Pythonic)
, addfind_package(Pythran)
+ everything related to its dependencies (i.e., Python, Numpy). I don't know much about pythran, but I assumed that before this PR thePythonic
dependency corresponded to the subpackage that is included in pythran. Is that right? At least I've successfully run all the benchmarks :-).